Signal on output grids even though master is fully off

[loogie]

Here's something I've been experimenting with.  I've built another and it works good, but this one gets a signal on the output grids even with the master fully CCW.  It gets louder as I turn VR1 up.  Its a strong, clean signal.  Its not going through the tone stack.  If I ground one of the grids of the PI it stops.  If I ground a PI grid through a cap it doesn't.  The signal isn't making it to the master.  I've chopsticked and followed the signal with the scope.  Otherwise it sounds great.

I'm sure its a wiring problem.  Or it could be some fatal flaw in the circuit.  I guess I'm wondering what I could look for?  Working schematic.  Apologies.

[eleventeen]

Casually glancing at the ckt, the "short" your master volume pot is attempting to throw into the ckt is not so short thru that .001 uf which is in series with it. Just guesstimating, that cap would have to be in the neighborhood of .1 uf (eg; 100x bigger) to let all guitar freqs thru and short them out. IMO you've created a low-pass filter and the lows are what's getting through. Not sure why you have that cap in there at all.

[sluckey]

but this one gets a signal on the output grids even with the master fully CCW.

Turn VR5 fully CCW too. Does that fix it?

Not sure why you have that cap in there at all.

That cap prevents the MV from interfering with the high impedance bootstrap bias used by the LTP PI.

[loogie]

Turn VR5 fully CCW too. Does that fix it?

If I turn VR5 fully CW it just about fixes it.  Still a faint sound.  Here's an updated schematic. 

[sluckey]

If I turn VR5 fully CW it just about fixes it.

Hmm. Maybe the outside lugs of VR5 are reversed? Usually you want the wiper to be grounded when a level or volume pot is turned max CCW.

OK, put a ground jumper on C13 and another on C17. Don't ground the tube side of the caps. That should totally kill any signal getting thru the PI to the PA grids. If not, then you have a wiring error, bad ground, or lead dress issue.

[loogie]

OK, put a ground jumper on C13 and another on C17. Don't ground the tube side of the caps. That should totally kill any signal getting thru the PI to the PA grids. If not, then you have a wiring error, bad ground, or lead dress issue.

Definitely a wiring problem.  I'll find it.  Thanks for your help.

The only other thing I can think of is bad filtering maybe because of a bad cap.  I seem to remember reading something in RDH4 about insufficient filtering across multiple stages.  I think it applies mostly across multiple to inverting stages where you can get positive feedback, but I can't remember. 

The idea behind the cut control is that it switches in a cap (c29) that cuts some of the highs in front and then as you turn it up it continues to cut at the PI.  The only pot with a switch I could find is a linear pot and for my application I think I need a reverse log.  Nothing much happens moving from CCW until you're very near CW and then it happens all at once.  I can make a reverse log with a resistor, but according my calculations a 1m pot with a 470k resistor would work best.  I could be wrong.

If I can't get satisfactory results I'll just use a DPDT like I did on my first prototype.  Everything I make is a prototype...